Scientific Method Scientific Method The primary goal o

  • Slides: 19
Download presentation
Scientific Method

Scientific Method

Scientific Method The primary goal o f science is to help us understand our

Scientific Method The primary goal o f science is to help us understand our universe. Knowledge is based on the gathering of reliable information. The objective (unbiased) means of gathering information is termed the scientific method. To accomplish this there are 4 basic steps 1. Observe 2. Describe. 3. Explain 4. Predict

 The primary means of observation is the scientific experiment. The primary purpose of

The primary means of observation is the scientific experiment. The primary purpose of the scientific experiment is to establish cause and effect. To do this one must create an experimental design that controls as many factors as possible with the exception of the ones you are studying. These factors are called variables. The variable that the scientist is studying is called the independent variable. It is the variable of interest. The one to which the study is trying to establish as being the cause of any observed differences.

The variable used to measure the effect of the independent variable is called the

The variable used to measure the effect of the independent variable is called the dependent variable. Other factors that may impact a change in performance are termed extraneous variables. Better research designs control more extraneous variables. Ideally all variable would be controlled except for the independent variable. In that case any difference in performance can then be solely attributed to the independent variable (i. e. the variable of interest) Two groups are compared one exposed to the independent variable (experimental group) and the other group that is not exposed to the independent variable (control group).

Illustration Experimental Group Independent Variable Control Group Extraneou s ? ? ? ? Extraneous

Illustration Experimental Group Independent Variable Control Group Extraneou s ? ? ? ? Extraneous ? ? ? Result -If the results are the same or not significantly different, then the Independent Variable is not effective, i. e. causes no significant difference -if there is a significant difference ask: Are there other variables (extraneous) that may have caused the difference. If all extraneous variable are controlled and are the same between the two groups then the difference can be attributable to the I. V.

Example A study of the effect of creatine supplementation on strength using a 1

Example A study of the effect of creatine supplementation on strength using a 1 RM bench press to measure strength Independent variable= ?

A study of the effect of creatine supplementation on strength using a 1 RM

A study of the effect of creatine supplementation on strength using a 1 RM bench press to measure strength Creatine is the independent variable. The Dependent variable is ?

A study of the effect of creatine supplementation on strength using a 1 RM

A study of the effect of creatine supplementation on strength using a 1 RM bench press to measure strength Dependent Variable = 1 RM bench press test Extraneous Variable = ?

A study of the effect of creatine supplementation on strength using a 1 RM

A study of the effect of creatine supplementation on strength using a 1 RM bench press to measure strength Extraneous Variables: Nutrition, motivation, level of fitness, sleep, age, etc.

Non-Causal Relationships/ Associations or Correlational Studies Simply try to establish a level of association

Non-Causal Relationships/ Associations or Correlational Studies Simply try to establish a level of association or correlation between two variables. No attempt to control variables or establish cause and effect r=. 95 =strong association between two variables Correlation statistic on 1. 0 equals a perfect correlation R=. 45 indicates a weak association

Correlation study example Correlation between vertical jump and I. Q. Correlation between lung cancer

Correlation study example Correlation between vertical jump and I. Q. Correlation between lung cancer and smoking Ethical problems to design a study to prove cause and effect. If correlation is high enough and ethics prevents verification of the hypothesis by the scientific method, and science can explain why there is the possibility of cause and effect, then it is generally accepted as truth ( or accepted theory) assuming a vast number of high correlations in many, many correlation studies Generally, correlation studies are used as a screen to see if a scientific experiment is warranted. If the correlation is low, then there would be a slim chance that there is a cause and effect present between the two variables. If it is high, then perhaps cause and effect can be established between two variables. Analysis is also required.

Types of Research Basic research- goal is directed to a theoretical generalization. Knowledge for

Types of Research Basic research- goal is directed to a theoretical generalization. Knowledge for knowledge’s sake. There may be no practical value in this knowledge e. g. the mating habits of the monarch butterfly, the effect of music on heart rate and perceived exertion, the effect of salt on the flow of water through a tube Applied research- goal is to solve a problem the researcher faces. e. g. how best to improve strength through training and nutrition in a short amount of time (8 weeks).

Theories A theory is an attempt to explain why we consistently observe a certain

Theories A theory is an attempt to explain why we consistently observe a certain effect or why an independent variable produces a certain effect. It is presented as an explanation and basis from which outcomes can be predicted related to situations in which the conditions addressed by theory have not yet been specifically tested. (creatine supplementation is shown to increase strength in bodybuilders, we can theorize it will do the same in unfit individuals)

Theories are constantly being tested and revised. Theories once accepted, can be rejected as

Theories are constantly being tested and revised. Theories once accepted, can be rejected as new evidence (observations) are discovered. Laws or facts are based on consistent results of many experiments (observations) that provide evidence that certain results can be consistently observed as the result of the influence of certain independent variables. (e. g. law of gravity, target heart rate)

Other Research Designs Case Studies-law, medicine, education Epidemiological Studies – data/statistical studies concerning disease

Other Research Designs Case Studies-law, medicine, education Epidemiological Studies – data/statistical studies concerning disease in a population Controlled Clinical Studies a. Blind- subjects do not know if they are in the treatment or control group b. Double Blind- Both the researcher and the subjects do not know who are in the treatment group. (Best design when many extraneous variable are controlled). Ofthen used in medicine and drug research.

Research Designs Cont’d Randomizes Subjects- helps to prevent bias. Samples are theorized to represent

Research Designs Cont’d Randomizes Subjects- helps to prevent bias. Samples are theorized to represent the entire population they must represent a cross-section of the population. (e. g. survey data on best cars in America, given to people in Detroit) Results from one population may not apply to another. (e. g. animal studies may not produce the same result in humans) (Anti-depressant drugs have different effects in adults vs. adolescents)

Other Controls for Quality/Unbiased Data Peer Review- research is reviewed by “experts” in the

Other Controls for Quality/Unbiased Data Peer Review- research is reviewed by “experts” in the field Peer Reviewed Journals-Journals in which research is not published until reviewed by tow or more experts (examining research design and conclusions) before being considered by the editor. Most professional and medical journals are peer reviewed. Meta-analysis- refers to a wide ranging review of research and the statistical analysis of it, averaging results and conclusions of many studies that address the research topic.

Bias in Research Publication Bias- Scientific Journals are more likely to publish positive rather

Bias in Research Publication Bias- Scientific Journals are more likely to publish positive rather than negative studies. Thus, there is a suppression of research that does not support drugs or ideas that are being favored or developed by companies, individuals, or institutions. (e. g. reading lab example). Conflicts of Interest: Financial Research funded by a drug company that fails to support their product may no longer be funded. (Sugar industry and saccharin)

Scientific Misconduct Scientists publish bogus data for financial gain e. g. the Lancet, 1998

Scientific Misconduct Scientists publish bogus data for financial gain e. g. the Lancet, 1998 (British Journal of Medicine) Dr. Wakefied had an article published in the Lancet and then suggested that the measles vaccine may be linked to autism a press conference. . Research did not show CAUSE & EFFECT (research was correlational) but Dr. W. vilified the vaccine in his press conference, inferring cause and effect. This was widely publicized on the WWW and the media and vaccination rates plummeted. 2010 - UK Medical Council (which licensed doctors) declared Dr. W. acted dishonestly and unethically. They retracted the article and Dr. W. ’s medical license was revoked by the UK Medical Council. (Hand out recent article on Autism)